Crusher79
Contributor
Contributor

Eternus DX100 S3 and ESXi 7 U2 only first device in LUN group working

Hi,

we are using for a special env. a

2x ProLiant DL560 Gen8
-- with HP SmartArary P822 (HBA Mode, v8.32)
Fujitsu Eternus DX100 S3
-- only available connection type: SAS
-- 2x LUN Groups with overlap
ESXi 7 U2
-- NMP

Default Eternus profile did not work. Device was listed as "Fujutsu Serial Attached..." instead of "Local Fujitsu...". Figured out that ALUA TPGS Mode does not work. After setting it to disable device was recognized correctly. We ran into claimdevice error and fixed it with a vendor rule - esxcli storage core claimrule add -r 49 -t vendor -V FUJITSU -M ETERNUS_DXL -P NMP --force-reserved.

After claimrule reload everything was fine. Shared LUN/ LUN Overlap works too.

Here is my question: We added additionnel LUN to each LUN Group. Only LUN 0 (First device in LUN group) is available in ESXi.

It is possible to present an Eternus DX100 S3 LUN Group with some LUNs in it to one or more ESXi host?

Default settings did not work at all. Had to change Host Response - disabled TPGS Mode. Maybe I missed some additional settings too.

Example:

LUN group 01 with LUNs: VOL01, VOL02, VOL03

Only first one - VOL01 - is presented to ESXi host. Limitations of SAS? Anyone got a clue?

Thx. greetz Maik

0 Kudos
6 Replies
aakalan
Enthusiast
Enthusiast

I don't think it's because a limitation of esxi. We have direct attached storage and we connected it to 2 different hosts.

have you done the cable connection to second host like in the attached file?

0 Kudos
Crusher79
Contributor
Contributor

Hi,

yes we did. Every host has a connection to each storage controllers.

We tried host groups with controller's sas ids and by replacing it with the response template. Doesn't make a differnt at all.

 

LUNesxi1 and LUNesxi2 are the groups. 1x volume per group - the same volume at all - so it is marked as "overlaped". As I mentioned before, we tried to put a among of LUNs into each group. But only the first one is presented to esxi. Really the first. If we delete it from the group, the next one is presented to ESXi.

We run into some problems with claimrules. Fixed it with a vendor-type one. I don't get it.

17-05-_2022_21-55-35.jpg

 

Attached a screenshot of current LUN groups. Look a little like a mess. We did some tryouts before. However, only Host LUN ID 0 is available at ESXi.

0 Kudos
aakalan
Enthusiast
Enthusiast

I see a fix for eternus in 6.7, sometimes these kind of fixes also applied to 7.0 but they are not putting it to fixed issues section. Is it possible you to upgrade your host to 7.0 U3d and try again?

  • PR 2710158: You must manually add the claim rules to an ESXi host for FUJITSU ETERNUS storage

    You must manually add the claim rules to an ESXi host for FUJITSU ETERNUS AB/HB Series storage arrays.

    This issue is resolved in this release. This fix sets Storage Array Type Plugin (SATP) to VMW_SATP_ALUA, Path Selection Policy (PSP) to VMW_PSP_RR, and Claim Options to tpgs_on as default for FUJITSU ETERNUS AB/HB Series storage arrays.

https://docs.vmware.com/en/VMware-vSphere/6.7/rn/esxi670-202103001.html

0 Kudos
Kenny31
Contributor
Contributor

Hello,

I have the same incident. I'm on the latest version of vmware 6.7 and I still can't get the other LUNs.

Do you still have the incident? @Crusher79 

thx

0 Kudos
maksym007
Hot Shot
Hot Shot

Have you updated such things as Drivers and Firmware, on both sides? I mean on ESXi hosts and Eternus?

0 Kudos
Kenny31
Contributor
Contributor

Yes for all

0 Kudos